The purpose of this project is to investigate the use of myoelectric signals as a control source for an existing hand-grasp neuroprosthesis system. The existing method of controlling is based on a chest-mounted, shoulder position transducer system. With the proposed device, control signals could be derived from weak myoelectric activity, making it possible for patients without sufficient shoulder mobility to use the neuroprosthesis. The emphasis of this project will be placed on attaining command signals that allow proportional control. Stimulus artifact will be removed from the myoelectric signal by signal blanking, band-pass filtering, and electrical isolation between stimulating and sensing circuits. The processed myoelectric signals will be interfaced to the control unit of the existing neuroprosthesis. A preliminary clinical study will be conducted with individuals having high level of tetraplegia and also with the users of the existing implanted hand system. We propose to use percutaneous intramuscular electrodes to sense myoelectric activity during Phase I of the project. These electrodes will be implanted by a hypodermic needle and can easily be withdrawn at the end of use. Feasibility of the project shall be determined by the performance of the prototype device. PROPOSED COMMERCIAL APPLICATIONS: The implantable device controlled by myoelectric signals can be prescribed to many patients with spinal cord injury. This new system will allow us to expand our market to patients with spinal cord lesions at C4 or higher levels. The existing external shoulder transducer will be removed; and the system will become more acceptable for the patients. The use of this device will lower the cost of attendant care and improve the quality of life of the patient.